Dynamic meshing for deformable image registration

Cai, Yiqi; Guo, Xiaohu; Zhong, Zichun; Mao, Weihua

doi:10.1016/j.cad.2014.08.009

Cited by 6 publications

(6 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All these existing superpixel methods heavily rely on the regular structure of pixels and cannot be directly applied to supervoxel segmentation of unorganized 3D point clouds. The proposed merge-swap framework for supervoxel generation was inspired by the optimization technique proposed by [20], which was used to efficiently build superpixels by [11,21]. Equipped with a tailored energy function for supervoxels and adapted merging and swapping operations for point clouds, the proposed method generates compact supervoxels that adhere well to object boundaries.…”

Section: Superpixel Generation From Rgb-d Imagesmentioning

confidence: 99%

Merge-Swap Optimization Framework for Supervoxel Generation from Three-Dimensional Point Clouds

et al. 2020

View full text Add to dashboard Cite

Surpervoxels are becoming increasingly popular in many point cloud processing applications. However, few methods have been devised specifically for generating compact supervoxels from unstructured three-dimensional (3D) point clouds. In this study, we aimed to generate high quality over-segmentation of point clouds. We propose a merge-swap optimization framework that solves any supervoxel generation problem formulated in energy minimization. In particular, we tailored an energy function that explicitly encourages regular and compact supervoxels with adaptive size control considering local geometric information of point clouds. We also provide two acceleration techniques to reduce the computational overhead. The performance of the proposed merge-swap optimization approach is superior to that of previous work in terms of thorough optimization, computational efficiency, and practical applicability to incorporating control of other properties of supervoxels. The experiments show that our approach produces supervoxels with better segmentation quality than two state-of-the-art methods on three public datasets.

show abstract

Section: Superpixel Generation From Rgb-d Imagesmentioning

confidence: 99%

Merge-Swap Optimization Framework for Supervoxel Generation from Three-Dimensional Point Clouds

et al. 2020

View full text Add to dashboard Cite

show abstract

“…These pixel swappings can be launched in an iterative scheme [CGZM15] until energy convergence. It is encouraged to swap pixels to neighboring cluster if the swapping decreases the energy.…”

Section: K-partition Optimizationmentioning

confidence: 99%

“…Swapping Parallelization: The swapping test of all boundary pixels in each iteration can be performed parallelly [CGZM15]. We adopt a GPU implementation to speed up this process.…”

Section: Implementation Detailsmentioning

confidence: 99%

“…From the perspective of energy optimization, there are two widely used approaches for MCVT computation: the Lloyd relaxation scheme [Llo82] and the quasi-Newton approach [LWL * 09]. Yet, our problem is in a discrete manner with pixel clustering, this optimization process can be greatly accelerated with a variational merging-swapping framework [CGZM15]. The merging step reduces the partition number from the number of pixels to k in a greedy Quadric Error Metric (QEM) fashion [GH97].…”

Section: K-partition Optimizationmentioning

confidence: 99%

“…It is encouraged to swap pixels to neighboring cluster if the swapping decreases the energy. These pixel swappings can be launched in an iterative scheme [CGZM15] until energy convergence. In each iteration, boundary pixels are tested on whether their swappings could possibly decrease the total energy, i.e., for a boundary pixel x ∊ 풞 i neighboring with 풞 j , the energy changes from the state of ( 풞 i , 풞 j ) to ( 풞 i ‐ x, 풞 j + x ).…”

Section: Image/video Superpixel Generationmentioning

confidence: 99%

See 2 more Smart Citations

Anisotropic Superpixel Generation Based on Mahalanobis Distance

Cai¹,

Guo²

2016

Computer Graphics Forum

Self Cite

View full text Add to dashboard Cite

Superpixels have been widely used as a preprocessing step in various computer vision tasks. Spatial compactness and color homogeneity are the two key factors determining the quality of the superpixel representation. In this paper, these two objectives are considered separately and anisotropic superpixels are generated to better adapt to local image content. We develop a unimodular Gaussian generative model to guide the color homogeneity within a superpixel by learning local pixel color variations. It turns out maximizing the log-likelihood of our generative model is equivalent to solving a Centroidal Voronoi Tessellation (CVT) problem. Moreover, we provide the theoretical guarantee that the CVT result is invariant to affine illumination change, which makes our anisotropic superpixel generation algorithm well suited for image/video analysis in varying illumination environment. The effectiveness of our method in image/video superpixel generation is demonstrated through the comparison with other state-of-the-art methods.

show abstract

Nonrigid Image Registration

Baggs

Tamir²

2017

Theory and Applications of Image Registration

View full text Add to dashboard Cite

An important hypothesis of the research reported in this paper is that dynamic space warping (DSW), a dynamic programming (DP) technique (Bellman 1965, Brown 1996), can be effectively used for object matching and for inference of the image registration warping function. For this end, a novel, robust, and efficient method for non-rigid image registration using DSW in the "length-code" domain is developed. Length codes are affine transformation invariant object shape features (Baggs 1996). DSW is utilized to estimate linear, nonlinear, and non-rigid warping (Keogh 2006). The DP algorithm maps length codes in an attempt to identify objects that appear in more than one image and uses their coordinates to register the images.

show abstract

Dynamic meshing for deformable image registration

Cited by 6 publications

References 43 publications

Merge-Swap Optimization Framework for Supervoxel Generation from Three-Dimensional Point Clouds

Merge-Swap Optimization Framework for Supervoxel Generation from Three-Dimensional Point Clouds

Anisotropic Superpixel Generation Based on Mahalanobis Distance

Nonrigid Image Registration

Contact Info

Product

Resources

About